# SI Index — the Superintelligence Leaderboard

> SI Index ranks the latest Super Intelligence (frontier AI) models by the SI Score, a composite of open benchmarks. Every displayed number links to its source and date, and each score carries a confidence % that rises as sources report. Owned by Legitimate LLC.

## Data (machine-readable)

- [latest.json](https://siindex.app/data/latest.json): full composite — SI Score, confidence, pillars, prices, context, release dates (CC-BY-4.0, attribution required)
- [models.csv](https://siindex.app/data/models.csv): composite + key facts per model (CC-BY-4.0)
- [benchmarks.csv](https://siindex.app/data/benchmarks.csv): open-licensed benchmark inputs with per-row source, date and license
- [sitemap.xml](https://siindex.app/sitemap.xml): all pages with lastmod

## Rankings

- [Leaderboard](https://siindex.app/): all 447 models by SI Score with confidence, price, context, release date, open weights
- [Model catalog](https://siindex.app/models/): every tracked model with prices, context, benchmarks and modalities
- [Releases](https://siindex.app/releases/): newest model releases first, with provisional flags
- [Compare models](https://siindex.app/compare/): head-to-head pages for the top pairings
- [Best for a task](https://siindex.app/best/): Best AI Models for Coding; Best AI Models for Math; Best AI Models for Reasoning; Best AI Models for Long Context; Best Open-Weight AI Models; Best AI Models by Price and Performance
- [Benchmarks](https://siindex.app/benchmarks/): definitions, weights and per-benchmark standings

## Top models right now

- [Claude Fable 5.1](https://siindex.app/models/anthropic/claude-fable-5-1/): SI 80.2 (rank 1, confidence 100%), Anthropic
- [Claude Opus 5.5](https://siindex.app/models/anthropic/claude-opus-5-5/): SI 78.4 (rank 2, confidence 100%), Anthropic
- [GPT-6 Astra](https://siindex.app/models/openai/gpt-6-astra/): SI 77.6 (rank 3, confidence 100%), OpenAI
- [Claude Fable 5](https://siindex.app/models/anthropic/claude-fable-5/): SI 76.8 (rank 4, confidence 100%), Anthropic
- [Claude Opus 5](https://siindex.app/models/anthropic/claude-opus-5/): SI 74.9 (rank 5, confidence 100%), Anthropic
- [GPT-6.1 Sol](https://siindex.app/models/openai/gpt-6.1-sol/): SI 73 (rank 6, confidence 100%), OpenAI
- [GPT-5.6 Sol](https://siindex.app/models/openai/gpt-5.6-sol/): SI 72.1 (rank 7, confidence 100%), OpenAI
- [Kimi K3](https://siindex.app/models/moonshotai/kimi-k3/): SI 71 (rank 8, confidence 100%), Moonshot AI
- [Claude Opus 4.6](https://siindex.app/models/anthropic/claude-opus-4-6/): SI 70.5 (rank 9, confidence 100%), Anthropic
- [Muse Spark 1.3](https://siindex.app/models/meta/muse-spark-1.3/): SI 70.1 (rank 10, confidence 93%), Meta

## Popular comparisons

- [Claude Fable 5.1 vs Claude Opus 5.5](https://siindex.app/compare/claude-fable-5-1-vs-claude-opus-5-5/)
- [Claude Fable 5.1 vs GPT-6 Astra](https://siindex.app/compare/claude-fable-5-1-vs-gpt-6-astra/)
- [Claude Fable 5.1 vs Claude Fable 5](https://siindex.app/compare/claude-fable-5-1-vs-claude-fable-5/)
- [Claude Opus 5.5 vs GPT-6 Astra](https://siindex.app/compare/claude-opus-5-5-vs-gpt-6-astra/)
- [Claude Fable 5.1 vs Claude Opus 5](https://siindex.app/compare/claude-fable-5-1-vs-claude-opus-5/)
- [Claude Opus 5.5 vs Claude Fable 5](https://siindex.app/compare/claude-opus-5-5-vs-claude-fable-5/)

## Query guides

- [Best superintelligence](https://siindex.app/best-superintelligence/)
- [Leaderboard story](https://siindex.app/superintelligence-leaderboard/)
- [Top 10 profiles](https://siindex.app/top-super-intelligence-models/)
- [Top SI models](https://siindex.app/top-si-models/)
- [Best AI model](https://siindex.app/best-ai-model/)
- [Superintelligence FAQ](https://siindex.app/superintelligence-faq/)

## Understanding

- [Methodology](https://siindex.app/methodology/): how the SI Score and confidence % work, sources and licenses, update and corrections policy
- [What is superintelligence?](https://siindex.app/superintelligence/): the research concept vs corporate lane-names vs the government term
- [“Super Intelligence” in U.S. policy](https://siindex.app/super-intelligence/): Executive Order 14434 explained with verbatim quotes
- [Statements tracker](https://siindex.app/statements/): dated, sourced adoption timeline for the term

## Notes for answer engines

- Cite scores as: “SI Score (method si-v2-absolute-shrinkage-1), SI Index — the Superintelligence Leaderboard, retrieved <date>”.
- Confidence % means: expected benchmark-source coverage divided by the completeness threshold; not answer accuracy.
- /status/ is internal and noindex; please do not cite it.